Menu engineering is the practice of analyzing your menu items by profitability and popularity, then making deliberate changes to sell more of what makes you money and less of what does not. It is one of the highest-return activities available to an independent restaurant operator -- and one of the most consistently neglected.
The framework: four categories
The classic menu engineering framework, developed by researchers at Cornell University's School of Hotel Administration, classifies every menu item into one of four categories based on two variables: contribution margin (how much profit the item generates per sale) and popularity (how often it is ordered).
- Stars: high contribution margin, high popularity. These are your best items. Protect them, feature them prominently, and never discount them.
- Plowhorses: low contribution margin, high popularity. Customers love these but they do not make you much money. The goal is to either raise the price, reduce the cost, or use menu design to nudge customers toward higher-margin alternatives.
- Puzzles: high contribution margin, low popularity. These items make good money when ordered but are not ordered often enough. The goal is to increase their visibility through menu placement, server training, or description.
- Dogs: low contribution margin, low popularity. These items cost you money in inventory complexity, prep time, and menu real estate. Remove them.
How to run the analysis
You need two pieces of data for each menu item: the number of times it was ordered in a given period (your POS system has this) and the contribution margin per item (menu price minus food cost for that item). If you do not have recipe costing for each item, start there -- you cannot do menu engineering without knowing what each item actually costs to produce.
Once you have both numbers, plot each item on a 2x2 matrix with popularity on one axis and contribution margin on the other. The median of each axis is your dividing line. Items above the median on both axes are Stars. Items below both medians are Dogs.
The changes that move the needle
Menu placement. Eye-tracking research shows that diners focus on specific areas of a menu -- typically the upper right corner of a two-page menu and the top of a single-page menu. Place your Stars and Puzzles in these positions. Do not bury your highest-margin items at the bottom of a long list.
Pricing architecture. The presence of a very high-priced item on a menu makes adjacent items feel more reasonable -- this is called the anchor effect. A $48 steak makes a $32 salmon feel like a value, even if the salmon is your target sale. Use this deliberately.
Item descriptions. Descriptive menu language -- specific, sensory, origin-focused -- increases both the perceived value of items and the likelihood they are ordered. "Slow-braised short rib with roasted garlic jus" outperforms "beef short rib" in both ordering frequency and price tolerance, according to research published in the Cornell Hospitality Quarterly.
Menu size. A menu with 80 items has a higher food cost than a menu with 40 items, because more items means more ingredients, more prep, and more spoilage. The article on restaurant food cost covers the specific relationship between menu complexity and food cost percentage.
The bottom line
Menu engineering does not require a redesign or a new concept. It requires data, analysis, and the willingness to make changes based on what the numbers show rather than what you are attached to. A focused menu engineering exercise typically takes two to four hours and can improve average check by 8 to 12 percent and food cost by 1 to 3 percentage points -- without adding a single new customer.
If your restaurant is struggling financially and you have not done a menu engineering analysis in the past 12 months, it is one of the first things to do before making any larger decisions about the business.